! git log -1 --format="%H"

28688145c190ebc1427f3830d5580e6337c2ebbf

import sys
# Install gurobi
!{sys.executable} -m pip install -i https://pypi.gurobi.com gurobipy
!{sys.executable} -m pip install plotly geopy

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import pandas as pd
from itertools import product

import gurobipy as gp
from gurobipy import GRB, quicksum

# These are ours
import problem
import visualization

Looking in indexes: https://pypi.gurobi.com
Requirement already satisfied: gurobipy in /opt/anaconda3/lib/python3.8/site-packages (9.1.1)
Requirement already satisfied: plotly in /opt/anaconda3/lib/python3.8/site-packages (4.14.3)
Requirement already satisfied: geopy in /opt/anaconda3/lib/python3.8/site-packages (2.1.0)
Requirement already satisfied: six in /opt/anaconda3/lib/python3.8/site-packages (from plotly) (1.15.0)
Requirement already satisfied: retrying>=1.3.3 in /opt/anaconda3/lib/python3.8/site-packages (from plotly) (1.3.3)
Requirement already satisfied: geographiclib<2,>=1.49 in /opt/anaconda3/lib/python3.8/site-packages (from geopy) (1.50)

def normalize(i, j):
    assert i != j
    return (i, j) if i < j else (j, i)

Problem configuration

from model import Autonomax, Config

# Run on the first |C| cities (mostly for testing)
C = 41

# What scenario we will be utilizing
scenario = 2
# The number of cities in the core net
NC = 3
# 1 if the core net should be a cycle; 0 if it shoul be a path.
Z = 1

autonomax = Autonomax(
    Config(
        cities=problem.cities[:C],
        distances=problem.D[:C, :C],
        demand=problem.B[scenario][:C],
        core_city_count=NC,
        core_net_is_cycle=Z,
    )
)

Academic license - for non-commercial use only - expires 2021-05-15
Using license file /Users/sjurwold/gurobi.lic

A = autonomax.model.getA()

count = lambda d: len(d) if isinstance(d, gp.tupledict) else 1

V = sum(count(v) for v in autonomax.variables.values())
C = sum(count(c) for c in autonomax.constraints.values())

print(f'V = {V}, C = {C}')

fig, ax = plt.subplots(1, figsize=(128, 128 * C/V))
ax.set_xticklabels([])
ax.set_yticklabels([])
plt.spy(A)


total = 0
print(f'\nCONSTRAINT SETS:')
for (name, constraint) in autonomax.constraints.items():
    constraints_in_set = count(constraint)
    print(f'{constraints_in_set:>5} | {name}')
    #plt.gcf().text(0.07, 0.69 - 0.4 * (total + 0.5 * count(constraint)) / C, name, fontsize=14)
    total += constraints_in_set
    plt.plot([0, V], [total - 0.5, total - 0.5], color='grey')
    
total = 0
print(f'\nVARIABLE SETS:')
for (name, variables) in autonomax.variables.items():
    variables_in_set = count(variables)
    print(f'{variables_in_set:>5} | {name}')
    #plt.gcf().text(0.1 + 0.8 * (total + 0.5*count(variables)) / V, 0.8, name, fontsize=14)
    total += count(variables)
    plt.plot([total - 0.5, total - 0.5], [0, C], color='grey')
    
plt.savefig('constraint-matrix.png', dpi=200)

V = 9389, C = 6153

CONSTRAINT SETS:
    1 | one_control_center
   41 | control_city_directly_connected
 1681 | reduce_control_center_symmetry
   41 | core_city_ub
    1 | cycle_or_path
   41 | disallow_core_tree
    1 | exactly_nc_core_cities
   41 | control_center_is_connected
 2460 | is_connectable
  123 | is_connected_timestep
   41 | connected_graph
   41 | conservation_of_flow
  820 | force_edge_if_flow
  820 | edge_cost_lb

VARIABLE SETS:
   41 | is_control_center
  820 | is_core_edge
   41 | is_core_city
  164 | is_connected_step
 5043 | is_connectable_step
  820 | flow
  820 | abs_flow
  820 | is_sub_edge
  820 | edge_cost

autonomax.model.Params.timeLimit = 1200.0
autonomax.model.optimize()

Changed value of parameter timeLimit to 1200.0
   Prev: inf  Min: 0.0  Max: inf  Default: inf
Gurobi Optimizer version 9.1.1 build v9.1.1rc0 (mac64)
Thread count: 8 physical cores, 8 logical processors, using up to 8 threads
Optimize a model with 6153 rows, 9389 columns and 30299 nonzeros
Model fingerprint: 0xc7b830a6
Model has 4 SOS constraints
Model has 820 general constraints
Variable types: 2460 continuous, 6929 integer (6929 binary)
Coefficient statistics:
  Matrix range     [1e+00, 3e+05]
  Objective range  [1e+00, 2e+04]
  Bounds range     [1e+00, 1e+02]
  RHS range        [7e-01, 4e+01]
Presolve removed 90 rows and 2633 columns
Presolve time: 0.10s
Presolved: 6063 rows, 6756 columns, 30429 nonzeros
Variable types: 2460 continuous, 4296 integer (4296 binary)
Found heuristic solution: objective 101614.81952

Root relaxation: objective 1.488799e+03, 5527 iterations, 0.16 seconds

    Nodes    |    Current Node    |     Objective Bounds      |     Work
 Expl Unexpl |  Obj  Depth IntInf | Incumbent    BestBd   Gap | It/Node Time

     0     0 1488.79878    0  153 101614.820 1488.79878  98.5%     -    0s
H    0     0                    92993.411858 1488.79878  98.4%     -    0s
     0     0 2055.13765    0  242 92993.4119 2055.13765  97.8%     -    0s
H    0     0                    28818.896409 2055.13765  92.9%     -    0s
     0     0 2950.69016    0  253 28818.8964 2950.69016  89.8%     -    0s
     0     0 2951.48088    0  228 28818.8964 2951.48088  89.8%     -    0s
     0     0 2951.84716    0  260 28818.8964 2951.84716  89.8%     -    0s
     0     0 3104.36147    0  319 28818.8964 3104.36147  89.2%     -    0s
     0     0 3111.90023    0  312 28818.8964 3111.90023  89.2%     -    1s
     0     0 3112.25235    0  311 28818.8964 3112.25235  89.2%     -    1s
     0     0 3123.97500    0  287 28818.8964 3123.97500  89.2%     -    1s
     0     0 3133.75015    0  294 28818.8964 3133.75015  89.1%     -    1s
     0     0 3134.42243    0  291 28818.8964 3134.42243  89.1%     -    1s
     0     0 3134.79246    0  285 28818.8964 3134.79246  89.1%     -    1s
     0     0 3141.65289    0  298 28818.8964 3141.65289  89.1%     -    1s
     0     0 3145.99541    0  271 28818.8964 3145.99541  89.1%     -    1s
     0     0 3146.33500    0  272 28818.8964 3146.33500  89.1%     -    1s
H    0     0                    25414.826158 3146.33500  87.6%     -    1s
     0     0 3147.72598    0  273 25414.8262 3147.72598  87.6%     -    1s
H    0     0                    24803.669498 3147.72598  87.3%     -    1s
     0     0 3148.29332    0  265 24803.6695 3148.29332  87.3%     -    1s
H    0     0                    23678.405041 3148.29332  86.7%     -    1s
     0     0 3150.68467    0  303 23678.4050 3150.68467  86.7%     -    1s
     0     0 3150.68467    0  272 23678.4050 3150.68467  86.7%     -    1s
H    0     0                    19410.589948 3150.68467  83.8%     -    2s
     0     2 3150.68900    0  259 19410.5899 3150.68900  83.8%     -    2s
H   68    75                    19347.398499 3345.60107  82.7%   363    3s
H   68    75                    18403.134936 3345.60107  81.8%   363    3s
H   69    75                    17507.500375 3345.60107  80.9%   360    3s
H   74    75                    17478.472099 3345.60107  80.9%   342    3s
H  309   274                    17437.061948 3345.60107  80.8%   163    4s
H  310   274                    17416.036632 3345.60107  80.8%   163    4s
   382   361 3659.87835   20  228 17416.0366 3345.60107  80.8%   159    5s
H  646   566                    17406.036633 3345.60107  80.8%   120    5s
H  666   534                    16762.313114 3345.60107  80.0%   117    5s
H 1030   822                    16751.178581 3345.60107  80.0%  94.1    6s
* 1116   799             172    15951.365184 3345.60107  79.0%  93.2    6s
* 1117   770             171    15431.365184 3345.60107  78.3%  93.2    6s
H 1235   846                    15430.782990 3345.60107  78.3%  91.1    6s
H 1237   838                    15418.705150 3345.60107  78.3%  91.8    6s
  1495   976 9876.64104   63  272 15418.7052 3345.60107  78.3%  83.9   10s
  1516   990 3673.06795   23  393 15418.7052 3371.99330  78.1%  82.7   15s
H 1524   944                    15418.705141 3414.88923  77.9%  82.3   16s
  1536   952 10087.3880   53  434 15418.7051 3474.27621  77.5%  81.7   20s
  1546   962 3557.35957   11  335 15418.7051 3509.89522  77.2%   122   29s
  1548   964 3655.49498   12  356 15418.7051 3655.49498  76.3%   124   31s
  1567   975 3863.13366   15  335 15418.7051 3825.87921  75.2%   137   35s
H 1664   969                    15418.705127 3864.00963  74.9%   163   39s
  1696   972 6731.71313   20  216 15418.7051 3864.00963  74.9%   165   40s
H 1715   929                    15418.705101 3864.00963  74.9%   167   41s
H 1717   884                    15418.705086 3864.00963  74.9%   168   41s
H 1722   840                    15418.705027 3864.00963  74.9%   169   41s
H 1803   799                    15418.705007 3864.00963  74.9%   180   43s
H 1805   761                    15418.704942 3864.00963  74.9%   180   43s
H 1807   739                    15418.704912 3864.00963  74.9%   180   44s
H 1812   703                    15418.704895 3864.00963  74.9%   181   44s
H 1813   670                    15318.705150 3864.00963  74.8%   181   44s
  1832   678 5231.99152   26  230 15318.7052 3864.00963  74.8%   182   45s
  2022   706 9312.92486   33  166 15318.7052 3864.00963  74.8%   199   50s
  2458   779 4797.56257   20  328 15318.7052 4029.75433  73.7%   209   55s
  2699   825 11518.5442   28  227 15318.7052 4029.75433  73.7%   216   60s
  3179   828     cutoff   21      15318.7052 5180.70055  66.2%   219   65s
H 3188   795                    15318.705148 5180.70055  66.2%   220   65s
H 3190   775                    15318.705136 5180.70055  66.2%   221   65s
  3589  1037     cutoff   42      15318.7051 6590.47848  57.0%   220   70s
  4115  1265 8352.81027   17  235 15318.7051 7546.09982  50.7%   219   75s
  4576  1431 10770.2011   32  199 15318.7051 7829.80665  48.9%   216   80s
  5046  1663 13666.2879   29  168 15318.7051 8196.28175  46.5%   215   85s
  5716  1948 12096.3314   45   83 15318.7051 8453.38674  44.8%   215   90s
H 6489  2197                    15318.705126 8833.18338  42.3%   213   95s
H 7264  2389                    15318.705113 9149.25368  40.3%   211  101s
  7724  2705 13358.6597   33  153 15318.7051 9227.54783  39.8%   210  105s
  8423  3015 13566.1238   36  161 15318.7051 9416.33800  38.5%   211  112s
  8746  3178 13335.1588   44  140 15318.7051 9459.98583  38.2%   211  146s
  9382  3451 10708.7657   36  149 15318.7051 9606.96927  37.3%   209  151s
 10039  3740     cutoff   26      15318.7051 9743.44588  36.4%   208  159s
 10458  3811 13313.5950   33  134 15318.7051 9788.99325  36.1%   207  162s
 10731  3989 11988.5665   25  233 15318.7051 9815.96389  35.9%   207  165s
 11232  4072 15166.9957   45   87 15318.7051 9935.75144  35.1%   206  201s
 11650  4209     cutoff   36      15318.7051 10011.1375  34.6%   206  205s
 12677  4494 10858.0746   36  140 15318.7051 10139.0016  33.8%   205  211s
 13268  4656 13371.0211   45   81 15318.7051 10233.7282  33.2%   204  215s
 14434  4976 14369.9911   27  211 15318.7051 10387.7703  32.2%   202  225s
 14892  5227 13086.3329   43   73 15318.7051 10439.9672  31.8%   203  231s
 15721  5388     cutoff   38      15318.7051 10557.7352  31.1%   201  236s
 16394  5415 13534.5849   48   68 15318.7051 10627.4053  30.6%   200  240s
 16580  5628 15074.7981   56  107 15318.7051 10668.7391  30.4%   201  245s
 17963  5852     cutoff   46      15318.7051 10786.9283  29.6%   201  253s
 18690  5948 12132.2311   38  174 15318.7051 10898.6313  28.9%   201  258s
 19138  5972     cutoff   43      15318.7051 10944.2638  28.6%   202  263s
 19881  6011     cutoff   49      15318.7051 11018.8667  28.1%   202  268s
 20601  6058 12569.1237   38  159 15318.7051 11103.3793  27.5%   202  277s
 21002  6143     cutoff   35      15318.7051 11162.2063  27.1%   203  283s
 21860  6152 13128.3460   47  102 15318.7051 11252.5368  26.5%   203  289s
 22688  6169 14710.1985   53   72 15318.7051 11335.8614  26.0%   203  295s
 23594  6184 14172.1980   43  120 15318.7051 11450.5059  25.3%   203  315s
 24136  6173 12917.1491   44   91 15318.7051 11466.4718  25.1%   203  386s
 25086  6215 13490.5242   35  184 15318.7051 11628.1974  24.1%   203  392s
 26009  6152 12040.4899   35  150 15318.7051 11685.8734  23.7%   203  416s
 26762  6083 infeasible   39      15318.7051 11803.4022  22.9%   204  422s
 27693  6036 12485.6578   42  120 15318.7051 11888.0053  22.4%   203  427s
 28287  5983 14112.8093   41  176 15318.7051 11946.5037  22.0%   203  432s
 29272  5896 14307.2075   51   58 15318.7051 12023.9874  21.5%   203  438s
 30205  5768     cutoff   34      15318.7051 12137.3658  20.8%   203  443s
 30981  5576 13169.6357   35  147 15318.7051 12215.5098  20.3%   203  449s
 31844  5383 infeasible   40      15318.7051 12335.2274  19.5%   202  454s
 32732  5188 13418.5839   40  132 15318.7051 12401.0767  19.0%   202  460s
 33594  5082     cutoff   45      15318.7051 12516.4779  18.3%   202  466s
 34569  4843 infeasible   22      15318.7051 12608.8860  17.7%   202  472s
 35461  4643     cutoff   43      15318.7051 12712.7714  17.0%   201  506s
 36397  4378     cutoff   41      15318.7051 12823.9244  16.3%   201  511s
 37300  4068 13751.2113   30  226 15318.7051 12937.5775  15.5%   200  599s
 38134  3738 14904.6970   39  137 15318.7051 13039.5602  14.9%   200  604s
 39030  3493     cutoff   40      15318.7051 13178.0679  14.0%   199  609s
 39516  3006     cutoff   49      15318.7051 13244.8704  13.5%   199  613s
 40563  2518 infeasible   52      15318.7051 13452.5191  12.2%   198  618s
 41565  2008     cutoff   39      15318.7051 13545.0783  11.6%   197  628s
 42571  1421 14134.6574   44  102 15318.7051 13747.2665  10.3%   196  633s
 43640   794     cutoff   40      15318.7051 13946.6184  8.96%   194  638s
 44467     0     cutoff   45      15318.7051 14104.2236  7.93%   193  642s

Cutting planes:
  Gomory: 32
  Cover: 1453
  Implied bound: 5
  Clique: 1
  MIR: 417
  StrongCG: 1
  Flow cover: 797
  GUB cover: 17
  Inf proof: 4
  Zero half: 33
  RLT: 203
  Relax-and-lift: 28

Explored 46324 nodes (8695438 simplex iterations) in 642.68 seconds
Thread count was 8 (of 8 available processors)

Solution count 10: 15318.7 15318.7 15318.7 ... 15430.8

Optimal solution found (tolerance 1.00e-04)
Best objective 1.531870511262e+04, best bound 1.531870511262e+04, gap 0.0000%

Solution

city_info = pd.DataFrame(autonomax.city_info())
city_info

	Index	Name	IsCoreCity	IsControlCenter	Demand	IngoingFlow	OutgoingFlow
0	0	Boden	False	False	2.3	2.800000e+00	5.1
1	1	Borås	False	False	1.9	2.520000e+01	27.1
2	2	Eskilstuna	False	False	1.3	5.800000e+00	7.1
3	3	Falun	False	False	1.9	1.394766e-12	1.9
4	4	Gävle	False	False	0.7	2.610000e+01	26.8
5	5	Göteborg	False	False	6.4	2.486900e-13	6.4
6	6	Halmstad	False	False	3.4	1.160000e+01	15.0
7	7	Haparanda	False	False	3.9	0.000000e+00	3.9
8	8	Helsingborg	False	False	3.7	7.900000e+00	11.6
9	9	Hudiksvall	False	False	1.3	2.230000e+01	23.6
10	10	Jönköping	True	False	2.9	1.144000e+02	117.3
11	11	Kalmar	False	False	2.8	1.058709e-12	2.8
12	12	Karlskrona	False	False	4.3	4.476419e-13	4.3
13	13	Karlstad	False	False	2.4	1.023182e-12	2.4
14	14	Kiruna	False	False	2.8	1.563194e-13	2.8
15	15	Kristianstad	False	False	4.5	3.623768e-13	4.5
16	16	Lidköping	False	False	0.7	8.100000e+00	8.8
17	17	Linköping	True	False	2.9	5.580000e+01	58.7
18	18	Luleå	False	False	4.4	9.000000e+00	13.4
19	19	Malmö	False	False	4.2	3.700000e+00	7.9
20	20	Motala	True	False	3.1	1.219000e+02	125.0
21	21	Norrköping	True	True	3.4	1.740000e+02	52.4
22	22	Nyköping	False	False	4.4	3.750000e+01	41.9
23	23	Sandviken	False	False	2.5	1.556089e-12	2.5
24	24	Skellefteå	False	False	1.4	1.340000e+01	14.8
25	25	Skövde	False	False	3.1	8.800000e+00	11.9
26	26	Stockholm	False	False	7.7	2.980000e+01	37.5
27	27	Sundsvall	False	False	2.1	2.020000e+01	22.3
28	28	Trelleborg	False	False	3.7	1.442402e-12	3.7
29	29	Uddevalla	False	False	4.0	1.257661e-12	4.0
30	30	Umeå	False	False	1.9	1.480000e+01	16.7
31	31	Uppsala	False	False	3.0	2.680000e+01	29.8
32	32	Varberg	False	False	3.8	1.500000e+01	18.8
33	33	Vetlanda	False	False	0.8	1.590000e+01	16.7
34	34	Vänersborg	False	False	4.1	4.000000e+00	8.1
35	35	Västervik	False	False	3.4	3.006484e-12	3.4
36	36	Västerås	False	False	3.9	1.900000e+00	5.8
37	37	Växjö	False	False	4.3	8.800000e+00	13.1
38	38	Örebro	False	False	2.2	2.400000e+00	4.6
39	39	Örnsköldsvik	False	False	2.5	1.670000e+01	19.2
40	40	Östersund	False	False	1.0	1.765699e-12	1.0

edge_info = pd.DataFrame(autonomax.edge_info())
edge_info

	From	To	Type	Flow	Cost	Distance
0	Skellefteå	Umeå	SUB	14.8	557.326132	111
1	Eskilstuna	Västerås	SUB	-5.8	61.716685	43
2	Gävle	Hudiksvall	SUB	-23.6	966.610215	118
3	Norrköping	Nyköping	SUB	-41.9	595.269663	58
4	Nyköping	Stockholm	SUB	-37.5	1022.500000	90
5	Linköping	Norrköping	CORE	-52.4	440.000000	44
6	Karlskrona	Växjö	SUB	4.3	150.077607	102
7	Motala	Örebro	SUB	-4.6	138.362916	92
8	Boden	Kiruna	SUB	-2.8	449.539100	291
9	Sundsvall	Östersund	SUB	-1.0	77.633542	166
10	Halmstad	Helsingborg	SUB	-11.6	267.571986	79
11	Jönköping	Vetlanda	SUB	-16.7	286.749284	65
12	Umeå	Örnsköldsvik	SUB	16.7	627.590973	111
13	Lidköping	Skövde	SUB	8.8	105.450180	49
14	Karlstad	Örebro	SUB	2.4	61.279919	77
15	Halmstad	Varberg	SUB	15.0	258.577201	65
16	Lidköping	Vänersborg	SUB	-8.1	129.045192	60
17	Uddevalla	Vänersborg	SUB	4.0	22.172747	21
18	Gävle	Sandviken	SUB	-2.5	19.882118	25
19	Hudiksvall	Sundsvall	SUB	-22.3	524.081992	81
20	Borås	Göteborg	SUB	-6.4	131.078630	71
21	Falun	Västerås	SUB	1.9	97.009877	128
22	Jönköping	Linköping	CORE	-58.7	1250.000000	125
23	Linköping	Västervik	SUB	-3.4	112.715626	97
24	Sundsvall	Örnsköldsvik	SUB	-19.2	878.974055	127
25	Jönköping	Motala	CORE	117.3	1080.000000	108
26	Vetlanda	Växjö	SUB	-13.1	263.087118	72
27	Borås	Varberg	SUB	-18.8	467.695643	84
28	Haparanda	Luleå	SUB	3.9	180.293133	124
29	Luleå	Skellefteå	SUB	13.4	659.953460	133
30	Stockholm	Uppsala	SUB	-29.8	550.119554	69
31	Motala	Norrköping	CORE	125.0	790.000000	79
32	Kristianstad	Växjö	SUB	4.5	197.061487	120
33	Gävle	Uppsala	SUB	26.8	403.877951	60
34	Helsingborg	Malmö	SUB	-7.9	126.105814	60
35	Borås	Jönköping	SUB	27.1	646.341239	82
36	Jönköping	Skövde	SUB	-11.9	284.330749	81
37	Boden	Luleå	SUB	5.1	39.194104	32
38	Malmö	Trelleborg	SUB	-3.7	33.196374	34
39	Eskilstuna	Norrköping	SUB	7.1	241.290923	102
40	Kalmar	Vetlanda	SUB	2.8	124.941927	119

core_cost = sum(edge_info[edge_info['Type'] == 'CORE']['Cost'])
subnet_cost = sum(edge_info[edge_info['Type'] == 'SUB']['Cost'])
total_cost = core_cost + subnet_cost

print(f'core   = {core_cost:>9.3f}')
print(f'subnet = {subnet_cost:>9.3f}')
print('------------------')
print(f'total  = {total_cost:>9.3f}')
assert abs(autonomax.model.getObjective().getValue() - total_cost) < 1e-6

core   =  3560.000
subnet = 11758.705
------------------
total  = 15318.705

visualization.show(pd.DataFrame(autonomax.city_info()), pd.DataFrame(autonomax.edge_info()))